lasio: Read/Write Well Data from LAS Files

0.32 · active · verified Fri Apr 17

The `lasio` library (version 0.32) provides robust tools for reading and writing well data from Log ASCII Standard (LAS) files, supporting both LAS 1.2 and 2.0 specifications. It is actively maintained with regular updates for bug fixes and feature enhancements, typically releasing new minor versions a few times a year.

Common errors

```
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xf1 in position ...: invalid start byte
```
cause The LAS file uses an encoding other than the default 'utf-8', such as 'latin-1' or 'cp1252', which cannot be decoded by 'utf-8'.

fix Specify the correct encoding when reading the file, e.g., `l = lasio.read('your_file.las', encoding='latin-1')`.
```
AttributeError: 'LASFile' object has no attribute 'well_name' (or similar for other direct well attributes)
```
cause Well information attributes (like WELL, COMP, DATE) are not directly accessible as properties of the `LASFile` object. They are stored within the `l.well` object.

fix Access well attributes through the `l.well` object and its `Item` objects, e.g., `well_name = l.well.WELL.value` or `company = l.well['COMP'].value`.
```
ValueError: No curve mnemonic found for curve with index ...
```
cause The LAS file contains data columns that do not have corresponding mnemonic definitions in the `~CURVE` section, indicating a malformed file.

fix Inspect the `~CURVE` section of the LAS file to ensure all data columns are properly defined with a unique mnemonic. If the file is uneditable, you might need to manually clean the data or skip the problematic section.

Warnings

gotcha LAS files frequently use non-standard encodings (e.g., 'latin-1', 'cp1252') that can cause `UnicodeDecodeError` if not specified. The default encoding for `lasio.read` is 'utf-8'.
Fix: Explicitly pass the correct encoding to `lasio.read()`. Common alternatives are `encoding='latin-1'` or `encoding='cp1252'` based on the file's origin.
breaking Version 0.20.0 introduced a significant refactor in how curve definitions and curve data are handled, using more explicit `CurveItem` objects. Direct access patterns used in older versions might break.
Fix: Review the 0.20.0 release notes. Ensure you are accessing curve metadata via `l.curves[idx].mnemonic`, `l.curves[idx].unit`, `l.curves[idx].value` and data via `l[mnemonic]` for consistency. Avoid direct mutation of `l.curves` lists if you previously did so.
gotcha When working with `lasio` data, remember to account for the `NULL` value defined in the LAS file header. `lasio` converts this value to `numpy.nan` by default.
Fix: Filter or impute `numpy.nan` values in your analysis, for example: `data_clean = l['GR'][~numpy.isnan(l['GR'])]` to remove NaNs before processing.

Install

pip install lasio Install stable version

Imports

lasio.read

import lasio
l = lasio.read('my_file.las')

lasio.LASFile
```
import lasio
l = lasio.LASFile()
```

Quickstart

This quickstart demonstrates how to read a LAS file (simulated from a string for simplicity) using `lasio.read()`, access well header information, retrieve curve data, and convert the data into a Pandas DataFrame.

import lasio
import io

# Simulate a LAS file content for demonstration
las_content = """~VERSION
VERS.      2.00:  LAS Version
WRAP.      NO:   No line wrapping
~WELL
STRT.M     10.00: START DEPTH
STOP.M     20.00: STOP DEPTH
STEP.M     0.50:  STEP
NULL.      -999.25: NULL VALUE
WELL.      EXAMPLE WELL: WELL
COMP.      EXAMPLE CO: COMPANY
~CURVE
DEPT.M     : Depth
GR.API     : Gamma Ray
NEUT.V/V   : Neutron
~A  DEPT      GR      NEUT
  10.000   100.0   0.300
  10.500   105.0   0.310
  11.000   110.0   0.320
  11.500   112.0   0.315
  12.000   115.0   0.325
"""

# Read the LAS file from a string (or from a path: lasio.read('path/to/file.las'))
las_file_obj = io.StringIO(las_content)
l = lasio.read(las_file_obj)

# Access well information
print(f"Well Name: {l.well.WELL.value}")
print(f"Company: {l.well.COMP.value}")

# Access curve data
print(f"Gamma Ray Curve (first 3 values): {l['GR'][:3].tolist()}")
print(f"Depth Curve (last 2 values): {l['DEPT'][-2:].tolist()}")

# Convert to Pandas DataFrame
df = l.df()
print("\nDataFrame head:")
print(df.head())

view raw JSON →